drm/amdgpu: Move umc ras block init to gmc ras sw_init

Initialize umc ras block only when umc ip block
supports ras. Driver queries ras capabilities after
early_init, ras block init needs to be moved to
sw_init.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Stanley Yang <Stanley.Yang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c

@@ -447,8 +447,15 @@ void amdgpu_gmc_filter_faults_remove(struct amdgpu_device *adev, uint64_t addr,
 	} while (fault->timestamp < tmp);
 }
 
-int amdgpu_gmc_ras_early_init(struct amdgpu_device *adev)
+int amdgpu_gmc_ras_sw_init(struct amdgpu_device *adev)
 {
+	int r;
+
+	/* umc ras block */
+	r = amdgpu_umc_ras_sw_init(adev);
+	if (r)
+		return r;
+
 	if (!adev->gmc.xgmi.connected_to_cpu) {
 		adev->gmc.xgmi.ras = &xgmi_ras;
 		amdgpu_ras_register_ras_block(adev, &adev->gmc.xgmi.ras->ras_block);

drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h

@@ -351,7 +351,7 @@ bool amdgpu_gmc_filter_faults(struct amdgpu_device *adev,
 			      uint16_t pasid, uint64_t timestamp);
 void amdgpu_gmc_filter_faults_remove(struct amdgpu_device *adev, uint64_t addr,
 				     uint16_t pasid);
-int amdgpu_gmc_ras_early_init(struct amdgpu_device *adev);
+int amdgpu_gmc_ras_sw_init(struct amdgpu_device *adev);
 int amdgpu_gmc_ras_late_init(struct amdgpu_device *adev);
 void amdgpu_gmc_ras_fini(struct amdgpu_device *adev);
 int amdgpu_gmc_allocate_vm_inv_eng(struct amdgpu_device *adev);

drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c

@@ -208,6 +208,36 @@ int amdgpu_umc_process_ras_data_cb(struct amdgpu_device *adev,
 	return amdgpu_umc_do_page_retirement(adev, ras_error_status, entry, true);
 }
 
+int amdgpu_umc_ras_sw_init(struct amdgpu_device *adev)
+{
+	int err;
+	struct amdgpu_umc_ras *ras;
+
+	if (!adev->umc.ras)
+		return 0;
+
+	ras = adev->umc.ras;
+
+	err = amdgpu_ras_register_ras_block(adev, &ras->ras_block);
+	if (err) {
+		dev_err(adev->dev, "Failed to register umc ras block!\n");
+		return err;
+	}
+
+	strcpy(adev->umc.ras->ras_block.ras_comm.name, "umc");
+	ras->ras_block.ras_comm.block = AMDGPU_RAS_BLOCK__UMC;
+	ras->ras_block.ras_comm.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
+	adev->umc.ras_if = &ras->ras_block.ras_comm;
+
+	if (!ras->ras_block.ras_late_init)
+		ras->ras_block.ras_late_init = amdgpu_umc_ras_late_init;
+
+	if (ras->ras_block.ras_cb)
+		ras->ras_block.ras_cb = amdgpu_umc_process_ras_data_cb;
+
+	return 0;
+}
+
 int amdgpu_umc_ras_late_init(struct amdgpu_device *adev, struct ras_common_if *ras_block)
 {
 	int r;

drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h

@@ -87,6 +87,7 @@ struct amdgpu_umc {
 	unsigned long active_mask;
 };
 
+int amdgpu_umc_ras_sw_init(struct amdgpu_device *adev);
 int amdgpu_umc_ras_late_init(struct amdgpu_device *adev, struct ras_common_if *ras_block);
 int amdgpu_umc_poison_handler(struct amdgpu_device *adev, bool reset);
 int amdgpu_umc_process_ecc_irq(struct amdgpu_device *adev,

drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c

@@ -699,25 +699,8 @@ static void gmc_v10_0_set_umc_funcs(struct amdgpu_device *adev)
 	default:
 		break;
 	}
-	if (adev->umc.ras) {
-		amdgpu_ras_register_ras_block(adev, &adev->umc.ras->ras_block);
-
-		strcpy(adev->umc.ras->ras_block.ras_comm.name, "umc");
-		adev->umc.ras->ras_block.ras_comm.block = AMDGPU_RAS_BLOCK__UMC;
-		adev->umc.ras->ras_block.ras_comm.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
-		adev->umc.ras_if = &adev->umc.ras->ras_block.ras_comm;
-
-		/* If don't define special ras_late_init function, use default ras_late_init */
-		if (!adev->umc.ras->ras_block.ras_late_init)
-				adev->umc.ras->ras_block.ras_late_init = amdgpu_umc_ras_late_init;
-
-		/* If not defined special ras_cb function, use default ras_cb */
-		if (!adev->umc.ras->ras_block.ras_cb)
-			adev->umc.ras->ras_block.ras_cb = amdgpu_umc_process_ras_data_cb;
-	}
 }
 
-
 static void gmc_v10_0_set_mmhub_funcs(struct amdgpu_device *adev)
 {
 	switch (adev->ip_versions[MMHUB_HWIP][0]) {
@@ -754,7 +737,6 @@ static void gmc_v10_0_set_gfxhub_funcs(struct amdgpu_device *adev)
 
 static int gmc_v10_0_early_init(void *handle)
 {
-	int r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	gmc_v10_0_set_mmhub_funcs(adev);
@@ -770,10 +752,6 @@ static int gmc_v10_0_early_init(void *handle)
 	adev->gmc.private_aperture_end =
 		adev->gmc.private_aperture_start + (4ULL << 30) - 1;
 
-	r = amdgpu_gmc_ras_early_init(adev);
-	if (r)
-		return r;
-
 	return 0;
 }
 
@@ -1024,6 +1002,10 @@ static int gmc_v10_0_sw_init(void *handle)
 
 	amdgpu_vm_manager_init(adev);
 
+	r = amdgpu_gmc_ras_sw_init(adev);
+	if (r)
+		return r;
+
 	return 0;
 }
 

drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c

@@ -581,23 +581,6 @@ static void gmc_v11_0_set_umc_funcs(struct amdgpu_device *adev)
 	default:
 		break;
 	}
-
-	if (adev->umc.ras) {
-		amdgpu_ras_register_ras_block(adev, &adev->umc.ras->ras_block);
-
-		strcpy(adev->umc.ras->ras_block.ras_comm.name, "umc");
-		adev->umc.ras->ras_block.ras_comm.block = AMDGPU_RAS_BLOCK__UMC;
-		adev->umc.ras->ras_block.ras_comm.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
-		adev->umc.ras_if = &adev->umc.ras->ras_block.ras_comm;
-
-		/* If don't define special ras_late_init function, use default ras_late_init */
-		if (!adev->umc.ras->ras_block.ras_late_init)
-			adev->umc.ras->ras_block.ras_late_init = amdgpu_umc_ras_late_init;
-
-		/* If not define special ras_cb function, use default ras_cb */
-		if (!adev->umc.ras->ras_block.ras_cb)
-			adev->umc.ras->ras_block.ras_cb = amdgpu_umc_process_ras_data_cb;
-	}
 }
 
 
@@ -846,6 +829,10 @@ static int gmc_v11_0_sw_init(void *handle)
 
 	amdgpu_vm_manager_init(adev);
 
+	r = amdgpu_gmc_ras_sw_init(adev);
+	if (r)
+		return r;
+
 	return 0;
 }
 

drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c

@@ -1318,23 +1318,6 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev)
 	default:
 		break;
 	}
-
-	if (adev->umc.ras) {
-		amdgpu_ras_register_ras_block(adev, &adev->umc.ras->ras_block);
-
-		strcpy(adev->umc.ras->ras_block.ras_comm.name, "umc");
-		adev->umc.ras->ras_block.ras_comm.block = AMDGPU_RAS_BLOCK__UMC;
-		adev->umc.ras->ras_block.ras_comm.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
-		adev->umc.ras_if = &adev->umc.ras->ras_block.ras_comm;
-
-		/* If don't define special ras_late_init function, use default ras_late_init */
-		if (!adev->umc.ras->ras_block.ras_late_init)
-				adev->umc.ras->ras_block.ras_late_init = amdgpu_umc_ras_late_init;
-
-		/* If not defined special ras_cb function, use default ras_cb */
-		if (!adev->umc.ras->ras_block.ras_cb)
-			adev->umc.ras->ras_block.ras_cb = amdgpu_umc_process_ras_data_cb;
-	}
 }
 
 static void gmc_v9_0_set_mmhub_funcs(struct amdgpu_device *adev)
@@ -1406,7 +1389,6 @@ static void gmc_v9_0_set_mca_funcs(struct amdgpu_device *adev)
 
 static int gmc_v9_0_early_init(void *handle)
 {
-	int r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* ARCT and VEGA20 don't have XGMI defined in their IP discovery tables */
@@ -1436,10 +1418,6 @@ static int gmc_v9_0_early_init(void *handle)
 	adev->gmc.private_aperture_end =
 		adev->gmc.private_aperture_start + (4ULL << 30) - 1;
 
-	r = amdgpu_gmc_ras_early_init(adev);
-	if (r)
-		return r;
-
 	return 0;
 }
 
@@ -1798,6 +1776,10 @@ static int gmc_v9_0_sw_init(void *handle)
 
 	gmc_v9_0_save_registers(adev);
 
+	r = amdgpu_gmc_ras_sw_init(adev);
+	if (r)
+		return r;
+
 	return 0;
 }